Abstract The goal of tfie proposed research is to determine mechanisnfis of viral pathogenesis using mammalian reovirus as an experimental system. Discrete steps in virus-host interaction often depend on selective recognition of cellular receptors and viral replication proteins that operate at distinct sites. However, the precise molecular mechanisms that govern viral dissemination and target cell selection are not well understood for most pathogenic viruses. This gap in knowledge has impeded the rational design of antiviral agents and vaccines that act at the earliest stages of infection. The proposed research uses reovirus, a highly tractable experimental model that shows promise for oncolytic and vaccine applications, to define how a virus interacts with its host at specific steps of the virus-host encounter. Reovirus infects a variety of mammalian species, including humans, but disease is restricted to the very young. Following peroral inoculation of newborn mice, reovirus infects intestinal mucosa, disseminates via hematogenous or neural routes, and targets the heart, liver, and central nervous system (CNS) to cause disease. Reovirus attachment protein a^ Influences the pathway of viral spread in the host and tropism for target tissues. The al protein is a filamentous trimer that binds to junctional adhesion molecule-A (JAM-A) and cell-surface sialic acid. In the previous funding period, we defined mechanisms by which a1 engages JAM-A and sialic acid, discovered that JAM-A-binding influences hematogenous spread and sialic acid-binding influences neural spread, and elucidated an essential function for nonstructural protein ols in reovirus dissemination. Our progress was facilitated by the development of a fully plasmid-based reverse genetics system for reovirus, the first of its kind for any mammalian double-stranded RNA virus. Based on these new discoveries, we hypothesize that reovirus spread to target tissues requires sequential recognition of discrete host-cell receptors by al and is dependent on ols. We propose three integrated specific aims in the extension period of this grant to test this hypothesis. In Specific Aim 1, mechanisms by which reovirus binding to JAM-A promotes infection of endothelial cells and hematogenous dissemination will be determined. We hypothesize that reovirus binding to JAM-A perturbs tight junction physiology in a manner to promote viral entry. To test this hypothesis, we will define the intracellular distribution of JAM-A and JAM-A-associated proteins following reovirus infection of polarized endothelial cells using confocal microscopy and subcellular fractionation. The phosphorylation status of JAM-A following reovirus attachment will be determined using wild-type and point-mutant JAM-A derivatives. Transgenic mice in which JAM-A is either expressed or ablated solely in endothelial or hematopoietic cells will be infected with reovirus and monitored for survival, viral load, and pathologic injury. These experiments will elucidate how a virus exploits a junction-associated receptor to initiate infection and clarify how a broadly expressed viral receptor dictates an exquisitely specific step in viral pathogenesis. In Specific Aim 2, receptors engaged by reovirus to infect the murine CNS will be identified. Although JAM-A is required for reovirus growth in endothelial cells, it is dispensable for reovirus growth in the brain. We hypothesize that reovirus uses unique receptors to infect the CNS. To test this hypothesis, we will identify neural receptors for reovirus using two complementary approaches. In the first, membrane preparations from JAM-A-null cortical neurons will be screened for candidate reovirus receptors using virus-overlay-protein- binding assays. In the second, a cDNA library will be generated from JAM-A-null brain tissue and screened for cDNAs that encode reovirus-binding molecules. Candidate receptors will be validated using receptor-specific antibodies, receptor-deficient cell lines, and assays of virus-receptor binding. The function of newly identified neural receptors in reovirus pathogenesis will be defined using receptor-null mice. These studies will reveal mechanisms by which viruses selectively target specific cells in the CNS. In Specific Aim 3, mechanisms by which the G1S protein influences reovirus pathogenesis will be elucidated. The als protein is required for reovirus dissemination and mediates cell-cycle arrest and apoptosis induction. We hypothesize that a1s-induced cell cycle arrest leads to apoptosis of target cells that are in turn engulfed by phagocytic ceils to initiate systemic dissemination. To test this hypothesis, we will identify sequences in a1s that influence reovirus cell cycle arrest, apoptosis induction, and systemic dissemination using mutant viruses engineered by reverse genetics. Targets of reovirus infection in vivo will be monitored for apoptotic cell death following inoculation with wild-type and als-mutant viruses to define the function of a1s- induced apoptosis in viral transit in the infected host. Cellular proteins engaged by a1s will be identified using tandem-affinity purification. Cells genetically deficient in candidate als-interacting proteins will be tested for reovirus infection, cell-cycle arrest, and apoptosis. These experiments will illuminate how a virus induces cell cycle arrest and apoptosis and the means by which these events lead to viral dissemination in the host. Results of the proposed experiments will contribute broadly to an understanding of mechanisms used by viruses to breach mucosal surfaces, disseminate systemically, and target specific host tissues to cause disease, The molecular blueprint of reovirus pathogenesis enabled by this work will identify key steps in virus- host interaction amenable to antiviral intervention and accelerate progress in the development of reovirus as a vector for oncolytic and vaccine applications.